Vacuum Chamber – Design and Build Advice

[blowlamp]

On blowlamp Said:

Whatever design you eventually choose, I think it would make sense to incorporate substantial external mounting points attached to and close to the pendulum suspension point, so as to reduce the possiblity of the structure rocking.

Hence why I’m asking.  My sums suggest a 3.7mm thick PVC pipe will withstand 0mb, whilst a 3.2mm thick pipe fails at about 170mm.

As no-one has experience of deliberately evacuating a PVC soil pipe, the purpose of the maths is to quantify the issue rather than just guessing.   As mentioned in the other thread I have successfully evacuated a smaller tube down to 500mb without any sign of stress.

My child-like maths gives me a surface area of about 196 square inches (14″ X 14″) of your PVC pipe, so when evacuated to -7.25 PSI, it seems that it is subject to 1421 pounds of pressure over the area. I’m not saying the pipe will implode, but if it’s not quite circular in cross section, then it might be vulnerable to further distortion over time or with a change in temperature.

All true, it might, so the purpose is to find how much.   If the answer is the pipe implodes at 800mb, I have to redesign it.   If it collapses at 500mb, I can usefully evacuate to 600mb, or apply hop.  If the maths says I can go all the way down to 0mb, then I can run with the maximum vacuum my pump can manage,

From the above, I think it follows that if you mount your clock purely on its base and some distortion does take place, it’s going to adversely affect performance.

Not unless it implodes!  And that definitely would ‘adversely affect performance’!

 

Martin.

What I don’t want to do is apply a lot of unnecessary countermeasures based on gut feel.  If the chamber cannot take the pressure, running the clock at more millibars is a better option, at least initially.

Dave

[Robert Atkinson 2]

On Robert Atkinson 2 Said:

On strengthing the plasitic cap with a metal disk. Stick the metal disk to the cap with a flexible sealant like Sikaflex EBT+ or RTV silcone. Then there will be no air presssure on the cap to make it bow…
If the disk is at least as big as the pipe OD and the plastic cap is fully seated forces will go vertically into the pipe wall.

Robert.

Makes sense.  It shall be so!

Ta,

Dave

[noel shelley]

On noel shelley Said:

I would advise against using PVC, in failure it may well shatter. A better choice is ABS pipe and for sealing a product called CT1 . Noel.

Good advice, except for the price:

Made a note though – nice thick walls.

There is a another worry plastics – it will gas under a vacuum.

Ta, Dave

[Robert Atkinson 2Participant @robertatkinson2]

The connector has about 35mm square flange and 28mm diameter body.
If the 7″ square base is big enough it looks ideal to me.
I’d send a picture but I’m not opening the garage in the current weather 🙁

Robert.

[Michael Gilligan]

On Michael Gilligan Said:

Whatever you choose to do, Dave

Avoid the use of brittle materials, and

Avoid any design details that risk sudden failure

I know, from bitter personal experience, that ‘catastrophic failure’ of a vacuum chamber is ‘non trivial’

MichaelG.

I can second these comments. I made a small vacuum chamber using circular Tupperware-style food container with a clip-on lid. It was obviously not an ideal vessel to use, the flat top and bottom aside. However, it worked well initially but, after a few uses, it imploded with an almighty bang. Fortunately, the shards of plastic did not go far, but it was a salutary lesson.

James.

[Robert Atkinson 2Participant @robertatkinson2]

Polycarbonate (PC) would be the safest choice for a clear tub. Not cheap though

https://www.clearplastictube.co.uk/polycarbonate-tube/clear-polycarbonate-tube-500mm-lengths/clear-polycarbonate-tube-extruded-250mm-5mm-500mm?sort=p.price&order=DESC

Tensile strength is about the same as acrylic but impat strength and thus tendancy to shatter is much higher. For a notched test a impact resistant acrylic is abou 1/20 of PC. PC has twice the notched impact strength of un-notched acrylic. This makes PC much less sensitive to damage.

There might even be an argument for a thinner lower strength acrylic or even glass tube inside a vented thin wall PC safety shield.

Robert.

[James A]

On James A Said:

On Michael Gilligan Said:

Whatever you choose to do, Dave

Avoid the use of brittle materials, and

Avoid any design details that risk sudden failure

I know, from bitter personal experience, that ‘catastrophic failure’ of a vacuum chamber is ‘non trivial’

MichaelG.

I can second these comments. I made a small vacuum chamber using circular Tupperware-style food container with a clip-on lid. It was obviously not an ideal vessel to use, the flat top and bottom aside. However, it worked well initially but, after a few uses, it imploded with an almighty bang. Fortunately, the shards of plastic did not go far, but it was a salutary lesson.

James.

Good points, so I looked the plastics up:

• Acrylic: very brittle with low impact strength.  Avoid!
• ABS:  moderately brittle,  3 – 7× tougher than PVC.  Pity it’s so expensive.
• PVC: brittleness depends on formulation, either side of ABS, but not as tough.  2× better than acrylic
• Polycarbonate.  Bullet-proof glass tough!  Unaffordium

In assessing the risk, I notice opinion ranges from NDIY being unimpressed by 500mb, to much more cautious recommendations. Be good to put numbers on it, so here’s question 3:

How much energy is contained by a 3.4 litre vacuum chamber?  And, as, the violence of an implosion is determined by how fast it progresses, how quickly would the energy be released?

In God we trust; all others bring data.

In the absence of data, my gut says the energy in the chamber is tiny compared with a model loco boiler filled with superheated water and steam pressurised at 10 bar.  But we know it is unwise to to test a boiler with compressed air.

The sums say the PVC pipe will fail by buckling, and I think the tube distorting will break the lower seal before the PVC suffers a brittle fracture.  (The pipe sits in a 4mm deep groove cut in the base, only held in place by weight and air pressure.)  I think it will fail with a slow pop, not a fast explosion.

For completeness, here’s the project’s risk assessment.  The vacuum chamber is a work in progress other than I’ve specified a 3× Safety Factor.  If necessary that can be achieved by running at a higher pressure.  A partial 800mb vacuum comfortably isolates the pendulum from atmospheric pressure and humidity change, and reduces friction.

Friction reduces linearly down to zero at 0mb, so 500mb reduces air resistance by half compared with atmospheric pressure.  A 400mb vacuum reduces air-resistance by 60%.    Lower vacuums are better, but there’s value in any reduction.

Dave

 

 

[John HaineParticipant @johnhaine32865]

Dave, I posted a link to Hall’s description of the Littlemore clock on the other thread.  I suggest you look at figure 3 in that article regarding the dependence of Q on pressure.

[SillyOldDuffer]

On SillyOldDuffer Said:

so here’s question 3:

<strong style=”font-family: Verdana, Arial, Helvetica, sans-serif; font-size: 14px;”>How much energy is contained by a 3.4 litre vacuum chamber?  And, as, the violence of an implosion is determined by how fast it progresses, how quickly would the energy be released?

In God we trust; all others bring data.

.

An interesting way to pose the question, Dave !

I would say the energy is outside not inside the chamber

But it’s the dynamics of the failure that matters … There is a near infinite supply of air-mass outside the chamber … Let 3.4 litres of it leak in gradually and nothing exciting happens … but a catastrophic failure is ‘instant’

My own incident occurred on a large ‘altitude chamber’ running at 100,00ft … and the destruction was immense.

I don’t really know how to scale that down to your sizes … but please heed my earlier advice.

MichaelG.

 

 

[Michael GilliganParticipant @michaelgilligan61133]

The Duck would have us believe that:

Search Assist

The mass of 1 liter of air is approximately 1.292 grams at standard temperature and pressure. This is based on the density of dry air, which is about 1.292 kg/m³.

Other estimates are available.

MichaelG.

[Michael Gilligan]

On Michael Gilligan Said:

The Duck would have us believe that:

Search Assist

The mass of 1 liter of air is approximately 1.292 grams at standard temperature and pressure. This is based on the density of dry air, which is about 1.292 kg/m³.

Other estimates are available.

MichaelG.

Mr Grok says, after showing the calculation:

344.5 J (or ≈0.096 Wh) of energy is released as the atmosphere crushes the tube. This is dissipated mainly as heat, sound, and plastic deformation of the material. (Real tubes buckle at lower ΔV\Delta VΔV, releasing less energy, but the query implies full collapse to zero volume.) 

Cause for concern rather than terrifying. Suitably applied 350J would lift my emaciated self 0.6m into the air.  Same as me falling 0.6m onto the floor.  No joke at my age.  Roughly equivalent to a 10km/h car crash. Unlikely to be fatal, but the risk has to be mitigated.

All this assumes the maths is right, and I’m famously bad at it.

🙁

Dave

[Wade BeattyParticipant @wadebeatty78296]

Could you use an alloy tube? Pieces can be found at the scrapyard or on ebay at much better prices

 

[Wade BeattyParticipant @wadebeatty78296]

And nothing would shatter in a failure

 

 

[Alan CharlestonParticipant @alancharleston78882]

Hi Dave,

I used to work in a lab where I would often  use an item of equipment called a vacuum desiccator. This consisted of a glass vessel with a flanged top edge. The top itself was also glass and had a flange on the bottom to mate with the vessel. It had a glass stopcock in the middle with a glass tube attached through which the vacuum was applied. The ones I used had a capacity of 5 – 10 litres.

I used a water aspirator vacuum pump to draw the vacuum. It is attached to a tap and water flows through a venturi which draws a vacuum through a side arm. From memory, the vacuum drawn was close to the theoretical value dictated by the vapour pressure of water. At 25 C this is 24mm of mercury which translates to a pressure of 32mB. You have to be careful using the aspirator – the chamber under vacuum needs to be isolated before turning the water off or water will be drawn into the chamber.

Regards,

Alan C.

[Robert Atkinson 2Participant @robertatkinson2]

The generally accepted energy value for pressure vessels, excluding steam, is 250 bar litre. so for vacuum anyhing unded 250l deos not require inspection or testing.
However brittle materials are to be avoided unless behind a safety shield. childhood experience of smashing old TV CRTs without implosion protection (drop lump hammer off flat shed roof at arms length) tells me big bits of glass will go to at leat 10ft vertically..

A potential issue with PVC, unlike the other plastics mentioned, is that it often contains plasticisers. These can leach out over time and cause the material to become brittle. Vacuum may speed this process.

Robert.

[John HaineParticipant @johnhaine32865]

+1 for metal tube – ebay has a number of sellers.

[BazyleParticipant @bazyle]

If worried by all the talk of the pipe shattering just remember the old wartime precaution of putting tape over the windows.

[SillyOldDuffer]

On SillyOldDuffer Said:

On Michael Gilligan Said:

The Duck would have us believe that:

Search Assist

The mass of 1 liter of air is approximately 1.292 grams at standard temperature and pressure. This is based on the density of dry air, which is about 1.292 kg/m³.

Other estimates are available.

MichaelG.

Mr Grok says, after showing the calculation:

344.5 J (or ≈0.096 Wh) of energy is released as the atmosphere crushes the tube. This is dissipated mainly as heat, sound, and plastic deformation of the material. (Real tubes buckle at lower ΔV\Delta VΔV, releasing less energy, but the query implies full collapse to zero volume.) 

Cause for concern rather than terrifying. Suitably applied 350J would lift my emaciated self 0.6m into the air.  Same as me falling 0.6m onto the floor.  No joke at my age.  Roughly equivalent to a 10km/h car crash. Unlikely to be fatal, but the risk has to be mitigated.

All this assumes the maths is right, and I’m famously bad at it.

🙁

Dave

I fear that Grok may have slightly missed the point … There is little, or no, buckling involved in a catastrophic failure of the vessel … just a very sudden inrush of air.

Another way of looking at it might be to assume that your pipe has a nett internal capacity of [say] 4g of dry air … That’s about 1/7th of a shotgun load, and I would posit that the velocity might be similar.

MichaelG.

.

[Robert Atkinson 2]

On Robert Atkinson 2 Said:

The connector has about 35mm square flange and 28mm diameter body.
If the 7″ square base is big enough it looks ideal to me.
I’d send a picture but I’m not opening the garage in the current weather 🙁

Robert.

I’m sure the connector is too big to fit into the Mk2 cast-iron base.   I’m reluctant to change the base at this stage, because the Mk2 clock isn’t operational yet, and I want to fully explore it’s potential and shortcomings before moving on.   The need for a bigger Mk3 clock is emerging for other reasons as well, and a 7″ base looks good for that.   Can I hold fire on it.  I’ve made a note in the Potential Improvements section of the design overview.  Kind offer, many thanks.

Dave

[noel shelleyParticipant @noelshelley55608]

What you seek may already exist in the form of a multi stage water pump – WITH OUT its guts ! Or an alloy tube, there is a source but I will not mention it in open forum. Dave contact me with a PM if the idea appeals. Noel.

[Michael Gilligan]

On Michael Gilligan Said:

On SillyOldDuffer Said:

On Michael Gilligan Said:

The Duck would have us believe that:

Search Assist

The mass of 1 liter of air is approximately 1.292 grams at standard temperature and pressure. This is based on the density of dry air, which is about 1.292 kg/m³.

Other estimates are available.

MichaelG.

Mr Grok says, after showing the calculation:

344.5 J (or ≈0.096 Wh) of energy is released as the atmosphere crushes the tube. This is dissipated mainly as heat, sound, and plastic deformation of the material. (Real tubes buckle at lower ΔV\Delta VΔV, releasing less energy, but the query implies full collapse to zero volume.) 

Cause for concern rather than terrifying. …

🙁

Dave

I fear that Grok may have slightly missed the point … There is little, or no, buckling involved in a catastrophic failure of the vessel … just a very sudden inrush of air.

I take the point, but the sums say it will fail by buckling, not in a catastrophic blink.  What I’d expect to happen is for the tube to bend inwards and in so doing pop itself out of the base recess, opening a large aperture in the container.

The sums suggest the pipe I have would support a 0mb vacuum with a safety factor a little below 2×.  In practice, I don’t have a pump capable of getting below 133mb, and it may be faulty.   Tested only down to 500mb.   The pressure sensor is on;y good down to 300mb, so I’m thinking 350-400mb ish in practice. That lifts the safety factor and reduces the energy.  I think it’s OL especially as the thing operates in isolation to reduce vibration.

…

The Titan submersible is at the other extreme.   It imploded in a few milliseconds with millions of tons of seawater pushing.  Instant death, just as well because the pressures would have compressed the air inside to spike over 1000°C, before the water cooled it an instant later.

Whilst I don’t think my clock is a widow-maker,I will be cautious.

On Robert Atkinson 2 Said:

…. childhood experience of smashing old TV CRTs without implosion protection (drop lump hammer off flat shed roof at arms length) tells me big bits of glass will go to at leat 10ft vertically..

Where we in the same gang?  Starting at age 15 my best friend made extra pocket money by mending TVs, on his own, and working for a TV repair man.  The latter had loads of old tubes for us to play with!  Lump hammer dropped on to the screen edge worked best, bang wise, breaking the neck was hopeless.  Never got glass to fly 10 feet.  Thinking about it, CRTs often failed by going soft, so maybe we didn’t get a decent implosion because they’d all lost some vacuum.  The getters were still silver though.

A potential issue with PVC, unlike the other plastics mentioned, is that it often contains plasticisers. These can leach out over time and cause the material to become brittle. Vacuum may speed this process.

Robert.

Mentioned in the Clock Measure thread. Over there the fear is that the plasticiser/solvents in the PVC will evaporate, drilling micro- tunnels through the wall.  In that scenario, PVC fails to hold a vacuum at all.  Here, I’m asking about the worst case, which is that the pipe does not leak, and is taken down to 0mb.

Other clocks:

• Shortt – metal tube, glass top.
• Livermore – glass.  (Glass is very good at resisting pressure applied slowly applied,  but it’s very brittle.)  The Livermore clock was located in a clock-house in the garden, sat on a buried 6 ton concrete pillar.
• Found a chap on the internet 2 years back using a bespoke Pyrex chamber.  He reported considerable difficulty degassing it.   Thought it was leaking, but it was gas entrapped during manufacture.   Took over a month to pump it out and hold the vacuum.   A deeper vacuum than I’m likely to achieve. I would never have guessed that Pyrex would gas, and it’s all too likely with PVC

A big advantage of PVC soil pipes is sealing the end is trivial: end-caps are off-the-shelf items.

Time will tell!

Dave

[John HaineParticipant @johnhaine32865]

LITTLEMORE!

[Michael GilliganParticipant @michaelgilligan61133]

I have evidently lost the plot again, Dave

… I thought you were seeking alternatives to your PVC pipe.

MichaelG.

[MacolmParticipant @macolm]

The quality of a calculation is proportional to the number of digits in the result.

For gas calculations, remember the skule rule, the volume of the gram molecular weight is 22.4 litres at standard temp and pressure.

For air, 20% oxygen, 80% nitrogen, near enough 29 combined molecular weight, this gives 1.2946428571428571428571428571429, clearly a better result than taking the opinion of some untrustworthy computer on the internet.

[duncan webster 1Participant @duncanwebster1]

Failure by buckling usually is in a blink, but in this case it can only go so far as the clock framework will at least impede it.

I once designed a much bigger vacuum vessel than this, it was not subject to pressure vessel regs as it wasn’t subject to internal pressure.

The maximum energy release is the volume × pressure, plus the strain energy in the tube, but without doing yet more sums I’d guess that is small in comparison

[Author]

Posts